Optimize with RAY¶

Home | Document | Github

Installation¶

At time of this tutorial, RAY support python <= 3.11, read more at Installation

In [ ]:

!pip install "ray[tune,client]"
!pip install optuna

!pip install "ray[tune,client]" !pip install optuna

Sample Strategy¶

In [1]:

import talib.abstract as ta

from lettrade import DataFeed, Strategy, indicator as i
from lettrade.exchange.backtest import ForexBackTestAccount, let_backtest

# from lettrade.indicator.vendor.qtpylib import inject_indicators

# inject_indicators()


class SmaCross(Strategy):
    ema1_period = 9
    ema2_period = 21

    def indicators(self, df: DataFeed):
        df["ema1"] = ta.EMA(df, timeperiod=self.ema1_period)
        df["ema2"] = ta.EMA(df, timeperiod=self.ema2_period)
        # df["ema1"] = df.close.ema(window=self.ema1_period)
        # df["ema2"] = df.close.ema(window=self.ema2_period)

        df["signal_ema_crossover"] = i.crossover(df.ema1, df.ema2)
        df["signal_ema_crossunder"] = i.crossunder(df.ema1, df.ema2)

    def next(self, df: DataFeed):
        if len(self.orders) > 0 or len(self.positions) > 0:
            return

        if df.l.signal_ema_crossover[-1]:
            price = df.l.close[-1]
            self.buy(size=0.1, sl=price - 0.001, tp=price + 0.001)
        elif df.l.signal_ema_crossunder[-1]:
            price = df.l.close[-1]
            self.sell(size=0.1, sl=price + 0.001, tp=price - 0.001)


lt = let_backtest(
    strategy=SmaCross,
    datas="example/data/data/EURUSD_5m-0_10000.csv",
    account=ForexBackTestAccount,
    # plotter=None,
)

import talib.abstract as ta from lettrade import DataFeed, Strategy, indicator as i from lettrade.exchange.backtest import ForexBackTestAccount, let_backtest # from lettrade.indicator.vendor.qtpylib import inject_indicators # inject_indicators() class SmaCross(Strategy): ema1_period = 9 ema2_period = 21 def indicators(self, df: DataFeed): df["ema1"] = ta.EMA(df, timeperiod=self.ema1_period) df["ema2"] = ta.EMA(df, timeperiod=self.ema2_period) # df["ema1"] = df.close.ema(window=self.ema1_period) # df["ema2"] = df.close.ema(window=self.ema2_period) df["signal_ema_crossover"] = i.crossover(df.ema1, df.ema2) df["signal_ema_crossunder"] = i.crossunder(df.ema1, df.ema2) def next(self, df: DataFeed): if len(self.orders) > 0 or len(self.positions) > 0: return if df.l.signal_ema_crossover[-1]: price = df.l.close[-1] self.buy(size=0.1, sl=price - 0.001, tp=price + 0.001) elif df.l.signal_ema_crossunder[-1]: price = df.l.close[-1] self.sell(size=0.1, sl=price + 0.001, tp=price - 0.001) lt = let_backtest( strategy=SmaCross, datas="example/data/data/EURUSD_5m-0_10000.csv", account=ForexBackTestAccount, # plotter=None, )

Optimize¶

In [2]:

from ray import tune, train
from ray.tune.search.optuna import OptunaSearch
from lettrade.exchange.backtest.extra.ray import LetOptimizeRay
import logging

logging.getLogger("ray.tune.tune").setLevel(logging.WARNING)

letray = LetOptimizeRay()
lettrade_model = lt.optimize_model(dumper=letray.dumps)


def train_model(config, data):
    params = {
        "ema1_period": config["ema1_period"],
        "ema2_period": config["ema2_period"],
    }

    # Model
    result = lettrade_model(params, optimizer_kwargs=data)

    # Score
    return {"score": result["equity"]}


search_space = {
    "ema1_period": tune.qrandint(5, 25, q=1),
    "ema2_period": tune.qrandint(10, 50, q=1),
}

algo = OptunaSearch()
# algo = tune.search.ConcurrencyLimiter(algo, max_concurrent=4)

tuner = tune.Tuner(
    tune.with_parameters(train_model, data=letray.data),
    tune_config=tune.TuneConfig(
        metric="score",
        mode="max",
        search_alg=algo,
        num_samples=100,
    ),
    # run_config=train.RunConfig(verbose=0),
    param_space=search_space,
)
results = tuner.fit()

lt.optimize_done()

from ray import tune, train from ray.tune.search.optuna import OptunaSearch from lettrade.exchange.backtest.extra.ray import LetOptimizeRay import logging logging.getLogger("ray.tune.tune").setLevel(logging.WARNING) letray = LetOptimizeRay() lettrade_model = lt.optimize_model(dumper=letray.dumps) def train_model(config, data): params = { "ema1_period": config["ema1_period"], "ema2_period": config["ema2_period"], } # Model result = lettrade_model(params, optimizer_kwargs=data) # Score return {"score": result["equity"]} search_space = { "ema1_period": tune.qrandint(5, 25, q=1), "ema2_period": tune.qrandint(10, 50, q=1), } algo = OptunaSearch() # algo = tune.search.ConcurrencyLimiter(algo, max_concurrent=4) tuner = tune.Tuner( tune.with_parameters(train_model, data=letray.data), tune_config=tune.TuneConfig( metric="score", mode="max", search_alg=algo, num_samples=100, ), # run_config=train.RunConfig(verbose=0), param_space=search_space, ) results = tuner.fit() lt.optimize_done()

Tune Status

Current time:	2024-06-15 09:31:24
Running for:	00:01:31.89
Memory:	9.4/62.7 GiB

System Info

Using FIFO scheduling algorithm.
Logical resource usage: 1.0/32 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:G)

Trial Status

Trial name	status	loc	ema1_period	ema2_period	iter	total time (s)	score
train_model_567e8489	TERMINATED	192.168.1.84:15993	17	14	1	0.394129	1056.08
train_model_0b9c709f	TERMINATED	192.168.1.84:16129	7	46	1	0.219453	920.68
train_model_15f951e4	TERMINATED	192.168.1.84:16203	18	14	1	0.187165	1116.68
train_model_1054961c	TERMINATED	192.168.1.84:16275	9	24	1	0.197212	893.48
train_model_aceac61d	TERMINATED	192.168.1.84:16341	24	38	1	0.205062	931.48
train_model_ce105e0c	TERMINATED	192.168.1.84:16413	23	17	1	0.183184	1020.28
train_model_a5ec8f89	TERMINATED	192.168.1.84:16511	20	41	1	0.206701	962.68
train_model_e1ead9ce	TERMINATED	192.168.1.84:16577	6	35	1	0.189952	960.68
train_model_9b0198ca	TERMINATED	192.168.1.84:16644	22	49	1	0.196741	1006.88
train_model_b22a89bc	TERMINATED	192.168.1.84:16711	5	24	1	0.19996	852.58
train_model_b0438800	TERMINATED	192.168.1.84:16782	8	21	1	0.196432	962.68
train_model_aa0e7c19	TERMINATED	192.168.1.84:16850	13	12	1	0.190239	1150.28
train_model_5efdbdc2	TERMINATED	192.168.1.84:16928	16	10	1	0.20991	1140.58
train_model_15871df1	TERMINATED	192.168.1.84:16996	12	11	1	0.201834	1153.88
train_model_3b28c8c8	TERMINATED	192.168.1.84:17063	13	10	1	0.200765	1153.78
train_model_9925dbec	TERMINATED	192.168.1.84:17129	13	29	1	0.189993	1029.18
train_model_b2764cd3	TERMINATED	192.168.1.84:17195	12	29	1	0.20041	979.58
train_model_15993c9d	TERMINATED	192.168.1.84:17261	11	19	1	0.225685	1012.48
train_model_ce4d58de	TERMINATED	192.168.1.84:17327	10	18	1	0.205284	950.48
train_model_2c7e1a4e	TERMINATED	192.168.1.84:17395	10	10	1	0.208917	1000
train_model_ba0b19e2	TERMINATED	192.168.1.84:17461	14	10	1	0.21732	1183.88
train_model_d04dbfda	TERMINATED	192.168.1.84:17547	14	25	1	0.254303	900.18
train_model_591bbf70	TERMINATED	192.168.1.84:17632	14	15	1	0.288824	992.68
train_model_c6e477f3	TERMINATED	192.168.1.84:17711	15	13	1	0.24678	1009.38
train_model_96bd7001	TERMINATED	192.168.1.84:17792	15	10	1	0.221637	1120.48
train_model_4d4d03cd	TERMINATED	192.168.1.84:17876	19	21	1	0.247008	969.68
train_model_72dbb93d	TERMINATED	192.168.1.84:17953	12	17	1	0.184375	992.78
train_model_ef05b46b	TERMINATED	192.168.1.84:18020	12	16	1	0.206489	980.68
train_model_49e81eb3	TERMINATED	192.168.1.84:18086	17	15	1	0.182067	1118.58
train_model_16349ae4	TERMINATED	192.168.1.84:18153	17	12	1	0.200432	1006.98
train_model_43578958	TERMINATED	192.168.1.84:18226	17	13	1	0.180715	1028.38
train_model_a74e8eb3	TERMINATED	192.168.1.84:18292	9	33	1	0.176146	899.68
train_model_03e41eec	TERMINATED	192.168.1.84:18358	13	10	1	0.172216	1153.78
train_model_fa885682	TERMINATED	192.168.1.84:18431	13	12	1	0.18103	1150.28
train_model_80f1f623	TERMINATED	192.168.1.84:18497	14	10	1	0.170597	1183.88
train_model_8e51adfb	TERMINATED	192.168.1.84:18564	11	10	1	0.191377	1177.58
train_model_30471d3f	TERMINATED	192.168.1.84:18657	15	20	1	0.185873	851.58
train_model_fd89af7d	TERMINATED	192.168.1.84:18723	11	20	1	0.1741	1031.88
train_model_912afd24	TERMINATED	192.168.1.84:18789	8	14	1	0.179255	810.48
train_model_60db4bd0	TERMINATED	192.168.1.84:18855	9	14	1	0.19334	856.08
train_model_81f9d953	TERMINATED	192.168.1.84:18921	19	41	1	0.180821	946.88
train_model_37878487	TERMINATED	192.168.1.84:18987	20	26	1	0.20723	941.18
train_model_e7b0b313	TERMINATED	192.168.1.84:19055	11	10	1	0.203764	1177.58
train_model_8298fcd9	TERMINATED	192.168.1.84:19121	11	11	1	0.192459	1000
train_model_a61dbe0a	TERMINATED	192.168.1.84:19193	11	12	1	0.239823	826.08
train_model_dbf37334	TERMINATED	192.168.1.84:19259	7	17	1	0.204537	846.08
train_model_bae0e248	TERMINATED	192.168.1.84:19325	7	16	1	0.184754	850.98
train_model_b6d4bc96	TERMINATED	192.168.1.84:19391	14	15	1	0.170364	992.68
train_model_0be6d479	TERMINATED	192.168.1.84:19457	14	50	1	0.168588	951.68
train_model_8ba14f88	TERMINATED	192.168.1.84:19523	16	13	1	0.187383	1007.38
train_model_4a7aa781	TERMINATED	192.168.1.84:19590	25	23	1	0.198682	1049.28
train_model_9d13b950	TERMINATED	192.168.1.84:19656	10	47	1	0.191259	912.48
train_model_05946552	TERMINATED	192.168.1.84:19722	10	10	1	0.183701	1000
train_model_bdc8be88	TERMINATED	192.168.1.84:19788	12	10	1	0.167025	1139.08
train_model_cd9ef040	TERMINATED	192.168.1.84:19854	12	11	1	0.171795	1153.88
train_model_0832acef	TERMINATED	192.168.1.84:19921	13	12	1	0.182617	1150.28
train_model_bcea84ba	TERMINATED	192.168.1.84:19990	13	12	1	0.176934	1150.28
train_model_8478f2b2	TERMINATED	192.168.1.84:20056	16	14	1	0.219377	1018.18
train_model_ffaab6d9	TERMINATED	192.168.1.84:20125	16	14	1	0.182941	1018.18
train_model_641cc62c	TERMINATED	192.168.1.84:20207	9	11	1	0.202532	901.88
train_model_3bd484ee	TERMINATED	192.168.1.84:20279	11	18	1	0.207438	992.38
train_model_77212f0f	TERMINATED	192.168.1.84:20346	12	18	1	0.197155	991.98
train_model_20deba80	TERMINATED	192.168.1.84:20414	12	11	1	0.197736	1153.88
train_model_263b7f9e	TERMINATED	192.168.1.84:20488	14	10	1	0.202125	1183.88
train_model_52395f8e	TERMINATED	192.168.1.84:20557	14	11	1	0.202452	1170.48
train_model_ed3c0fdd	TERMINATED	192.168.1.84:20627	14	13	1	0.194707	1059.38
train_model_482e6539	TERMINATED	192.168.1.84:20693	14	13	1	0.191817	1059.38
train_model_99a3604a	TERMINATED	192.168.1.84:20766	15	16	1	0.188537	944.08
train_model_9962248d	TERMINATED	192.168.1.84:20832	15	16	1	0.201938	944.08
train_model_39a0dea5	TERMINATED	192.168.1.84:20904	10	10	1	0.223073	1000
train_model_13d09268	TERMINATED	192.168.1.84:20972	13	35	1	0.200546	1001.28
train_model_c9fc7cd6	TERMINATED	192.168.1.84:21040	13	42	1	0.197765	913.68
train_model_ebaa73f9	TERMINATED	192.168.1.84:21106	12	11	1	0.216298	1153.88
train_model_c89fd048	TERMINATED	192.168.1.84:21175	11	11	1	0.205003	1000
train_model_210abe74	TERMINATED	192.168.1.84:21258	11	11	1	0.194978	1000
train_model_19a22c6e	TERMINATED	192.168.1.84:21328	14	12	1	0.192942	1060.38
train_model_3ecca919	TERMINATED	192.168.1.84:21394	14	13	1	0.206201	1059.38
train_model_8e85dfc0	TERMINATED	192.168.1.84:21494	15	15	1	0.219632	1000
train_model_5601e16a	TERMINATED	192.168.1.84:21561	12	15	1	0.183326	960.38
train_model_4b260a19	TERMINATED	192.168.1.84:21631	12	10	1	0.186971	1139.08
train_model_258cb360	TERMINATED	192.168.1.84:21697	16	10	1	0.197186	1140.58
train_model_6c38c970	TERMINATED	192.168.1.84:21764	13	10	1	0.188838	1153.78
train_model_fccdeabd	TERMINATED	192.168.1.84:21830	12	11	1	0.180142	1153.88
train_model_3bb0fd66	TERMINATED	192.168.1.84:21896	12	12	1	0.218196	1000
train_model_7789984e	TERMINATED	192.168.1.84:21962	10	13	1	0.182172	826.18
train_model_9970c744	TERMINATED	192.168.1.84:22029	10	13	1	0.204211	826.18
train_model_a3055312	TERMINATED	192.168.1.84:22095	9	11	1	0.200074	901.88
train_model_abab1cb7	TERMINATED	192.168.1.84:22167	22	14	1	0.205525	1158.88
train_model_369f6fab	TERMINATED	192.168.1.84:22270	13	28	1	0.190361	1000.18
train_model_d7d815b3	TERMINATED	192.168.1.84:22336	18	14	1	0.194311	1116.68
train_model_9940ae96	TERMINATED	192.168.1.84:22403	22	32	1	0.194089	923.28
train_model_b6a41835	TERMINATED	192.168.1.84:22469	22	12	1	0.183299	1168.28
train_model_38cbd8db	TERMINATED	192.168.1.84:22535	15	12	1	0.190919	1039.58
train_model_9e2cb1ef	TERMINATED	192.168.1.84:22601	22	12	1	0.209193	1168.28
train_model_381afd88	TERMINATED	192.168.1.84:22667	23	11	1	0.20926	1128.28
train_model_b587a096	TERMINATED	192.168.1.84:22733	23	14	1	0.183836	1100.68
train_model_0db4c953	TERMINATED	192.168.1.84:22801	21	14	1	0.196457	1148.58
train_model_a6b16ae2	TERMINATED	192.168.1.84:22867	21	10	1	0.203772	1007.98
train_model_ce670b25	TERMINATED	192.168.1.84:22935	22	12	1	0.21155	1168.28
train_model_b9da7392	TERMINATED	192.168.1.84:23002	24	12	1	0.197647	1168.98

Plot¶

In [3]:

lt.plotter.heatmap(x="ema1_period", y="ema2_period", z="equity")

lt.plotter.heatmap(x="ema1_period", y="ema2_period", z="equity")

In [4]:

lt.plotter.contour(x="ema1_period", y="ema2_period", z="equity")

lt.plotter.contour(x="ema1_period", y="ema2_period", z="equity")

Plot plotly¶

Tutorial

Init Plotly environment¶

In [5]:

import plotly.io as pio

pio.renderers.default = "notebook"
pio.templates.default = "plotly_dark"

import plotly.io as pio pio.renderers.default = "notebook" pio.templates.default = "plotly_dark"

In [6]:

df = results.get_dataframe()
df

df = results.get_dataframe() df

Out[6]:

	score	timestamp	checkpoint_dir_name	done	training_iteration	trial_id	date	time_this_iter_s	time_total_s	pid	hostname	node_ip	time_since_restore	iterations_since_restore	config/ema1_period	config/ema2_period	logdir
0	1056.08	1718418594	None	False	1	567e8489	2024-06-15_09-29-54	0.394129	0.394129	15993	pc	192.168.1.84	0.394129	1	17	14	567e8489
1	920.68	1718418594	None	False	1	0b9c709f	2024-06-15_09-29-54	0.219453	0.219453	16129	pc	192.168.1.84	0.219453	1	7	46	0b9c709f
2	1116.68	1718418595	None	False	1	15f951e4	2024-06-15_09-29-55	0.187165	0.187165	16203	pc	192.168.1.84	0.187165	1	18	14	15f951e4
3	893.48	1718418596	None	False	1	1054961c	2024-06-15_09-29-56	0.197212	0.197212	16275	pc	192.168.1.84	0.197212	1	9	24	1054961c
4	931.48	1718418597	None	False	1	aceac61d	2024-06-15_09-29-57	0.205062	0.205062	16341	pc	192.168.1.84	0.205062	1	24	38	aceac61d
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
95	1100.68	1718418680	None	False	1	b587a096	2024-06-15_09-31-20	0.183836	0.183836	22733	pc	192.168.1.84	0.183836	1	23	14	b587a096
96	1148.58	1718418681	None	False	1	0db4c953	2024-06-15_09-31-21	0.196457	0.196457	22801	pc	192.168.1.84	0.196457	1	21	14	0db4c953
97	1007.98	1718418682	None	False	1	a6b16ae2	2024-06-15_09-31-22	0.203772	0.203772	22867	pc	192.168.1.84	0.203772	1	21	10	a6b16ae2
98	1168.28	1718418683	None	False	1	ce670b25	2024-06-15_09-31-23	0.211550	0.211550	22935	pc	192.168.1.84	0.211550	1	22	12	ce670b25
99	1168.98	1718418684	None	False	1	b9da7392	2024-06-15_09-31-24	0.197647	0.197647	23002	pc	192.168.1.84	0.197647	1	24	12	b9da7392

100 rows × 17 columns

In [7]:

import pandas as pd

df.rename(
    columns={
        "config/ema1_period": "ema1_period",
        "config/ema2_period": "ema2_period",
    },
    inplace=True,
)
df

import pandas as pd df.rename( columns={ "config/ema1_period": "ema1_period", "config/ema2_period": "ema2_period", }, inplace=True, ) df

Out[7]:

	score	timestamp	checkpoint_dir_name	done	training_iteration	trial_id	date	time_this_iter_s	time_total_s	pid	hostname	node_ip	time_since_restore	iterations_since_restore	ema1_period	ema2_period	logdir
0	1056.08	1718418594	None	False	1	567e8489	2024-06-15_09-29-54	0.394129	0.394129	15993	pc	192.168.1.84	0.394129	1	17	14	567e8489
1	920.68	1718418594	None	False	1	0b9c709f	2024-06-15_09-29-54	0.219453	0.219453	16129	pc	192.168.1.84	0.219453	1	7	46	0b9c709f
2	1116.68	1718418595	None	False	1	15f951e4	2024-06-15_09-29-55	0.187165	0.187165	16203	pc	192.168.1.84	0.187165	1	18	14	15f951e4
3	893.48	1718418596	None	False	1	1054961c	2024-06-15_09-29-56	0.197212	0.197212	16275	pc	192.168.1.84	0.197212	1	9	24	1054961c
4	931.48	1718418597	None	False	1	aceac61d	2024-06-15_09-29-57	0.205062	0.205062	16341	pc	192.168.1.84	0.205062	1	24	38	aceac61d
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
95	1100.68	1718418680	None	False	1	b587a096	2024-06-15_09-31-20	0.183836	0.183836	22733	pc	192.168.1.84	0.183836	1	23	14	b587a096
96	1148.58	1718418681	None	False	1	0db4c953	2024-06-15_09-31-21	0.196457	0.196457	22801	pc	192.168.1.84	0.196457	1	21	14	0db4c953
97	1007.98	1718418682	None	False	1	a6b16ae2	2024-06-15_09-31-22	0.203772	0.203772	22867	pc	192.168.1.84	0.203772	1	21	10	a6b16ae2
98	1168.28	1718418683	None	False	1	ce670b25	2024-06-15_09-31-23	0.211550	0.211550	22935	pc	192.168.1.84	0.211550	1	22	12	ce670b25
99	1168.98	1718418684	None	False	1	b9da7392	2024-06-15_09-31-24	0.197647	0.197647	23002	pc	192.168.1.84	0.197647	1	24	12	b9da7392

100 rows × 17 columns

Type 1¶

In [8]:

from plotly import express as px

fig = px.scatter(df, x=df.index, y="score")
fig.show()

from plotly import express as px fig = px.scatter(df, x=df.index, y="score") fig.show()

Type 2¶

In [9]:

import plotly.express as px


fig = px.density_contour(
    df,
    x="ema1_period",
    y="ema2_period",
    z="score",
    histfunc="max",
)
fig.update_traces(contours_coloring="fill", contours_showlabels=True)
fig.show()

import plotly.express as px fig = px.density_contour( df, x="ema1_period", y="ema2_period", z="score", histfunc="max", ) fig.update_traces(contours_coloring="fill", contours_showlabels=True) fig.show()

Type 3¶

In [10]:

import plotly.express as px

fig = px.density_heatmap(
    df,
    x="ema1_period",
    y="ema2_period",
    z="score",
    nbinsx=20,
    nbinsy=40,
    histfunc="max",
    color_continuous_scale="Viridis",
)
fig.show()

import plotly.express as px fig = px.density_heatmap( df, x="ema1_period", y="ema2_period", z="score", nbinsx=20, nbinsy=40, histfunc="max", color_continuous_scale="Viridis", ) fig.show()